This invention relates generally to electric power grids and more particularly, to methods and apparatus for operating combined-cycle power systems.
The term “full load” is used herein interchangeably with “rated output” and “maximum continuous rating” (MCR). These terms refer to an upper range of continuous operation output for the power system and its associated components. “Partial load” refers to an output level below full load.
Electric power grids typically include a number of power generating systems to supply electricity to the grid and a number of electricity consumers that draw electricity from the grid. When the generation and consumption of electricity are substantially equal, the grid frequency is substantially constant. Grid frequency is normally a parameter maintained at a substantially stable value. Examples of nominal standard grid frequencies for the European and North American systems are 50 Hz and 60 Hz, respectively.
Frequency deviations of a transient nature may result from increased or decreased consumption and/or removal or addition of power generation systems. Increased consumption and removal of power generation systems tends to cause a decrease of the grid frequency. Decreased consumption and addition of power generation systems tends to cause an increase of the grid frequency. Power consumption and power generation are time-dependent variables which may cause frequency variations in a range of approximately +0.5 Hz to −0.5 Hz. Generally, frequency transients are of a short duration, i.e., measured in seconds to minutes, and as discussed above, small magnitudes. The magnitude of a frequency transient is typically influenced by a ratio of the magnitude of a power variation to the total power level within the grid and associated interconnected grids throughout the duration of the variation. The aforementioned small magnitude frequency transients are consistent with the small size of a typical power variation as compared to the typically large size of nominal interconnected grids. Also, in general, power grids tend to be self-correcting with respect to maintaining grid frequency within a substantially constant range. For example, in the event of a frequency deviation from a standard value, a near-term variation in power generation spread over a number of power generator systems may be facilitated by at least one control system and at least one control strategy to mitigate the magnitude and the duration of the frequency transient such that frequency transients normally do not impact consumers.
Larger frequency transients such as transients greater than approximately +0.5 Hz to −0.5 Hz and due, for example, to a frequency decrease as a result of an immediate loss of one or more power generators, sometimes referred to as a trip, may tend to induce a large frequency decrease. One possible method to mitigate the frequency transient magnitude and duration is to have some amount of standby power generation capacity, sometimes referred to as system reserve, available within the grid to respond to the frequency decrease within seconds of the transient. For example, a particular generating unit on the grid may be induced to initiate a fast increase in its associated power generation output to the grid.
Many known power generation facilities include either steam turbine generators (STG), combustion turbine generators (CTG), or some combination thereof. These configurations typically include a turbine rotatably coupled to an associated electric generator. The generator frequency is normally synchronized to the electric power grid frequency and rotates at a speed substantially similar to the grid frequency.
Many known STGs operate in flow communication with a steam generation apparatus, for example, a boiler. Generally, air and fuel are combusted to release thermal energy that is subsequently used to boil water to generate steam. The steam generated is channeled to a turbine wherein the thermal energy of the steam is converted to mechanical energy to rotate the rotor of the turbine. The power generated is proportional to the rate of steam flow to the turbine.
One known method of maintaining a power reserve is to operate a STG with at least one associated steam supply control valve in a partially open, or throttled, position such that the steam generator, the STG and the power grid are in an equilibrium, sometimes referred to as a steady-state condition, operating at some value less than full rated load of the steam generator and STG arrangement. The difference between full load and partial load is often referred to as spinning reserve. A controller is utilized to sense a decrease in system frequency and to generate a control signal transmitted to the steam valve within seconds of sensing a frequency transient. The control signal causes the valve to move to a more open position and thermal energy stored within the components of the steam generation apparatus, for example, the superheater, begins to be removed immediately via increased steam flow through the steam generator. Cooling fluid, air and fuel are subsequently increased over time to facilitate establishing a modified equilibrium between the steam generator, the STG and the power grid. However, many steam generator and STG combinations may take two to five minutes to attain the modified equilibrium while operating within predetermined parameters to mitigate the potential for increased stress and wear on affected components. Also, the amount of thermal energy typically stored in the aforementioned manner is limited. In addition, many steam generator and STG combinations may not effectively respond to a grid frequency transient with a stable, controlled response. For example, the aforementioned steam valve to the STG may open too quickly and deplete the thermal energy reserve too rapidly to deliver a sustained, effective response. Alternatively, the steam valve to the STG may open too slowly to deliver a timely, effective response.
Many known CTGs ignite a fuel-air mixture in a combustor assembly and generate a combustion gas stream that is channeled to a turbine assembly via a hot gas path. Compressed air is channeled to the combustor assembly by a compressor assembly that is normally coupled to the turbine, i.e., the compressor, turbine and generator rotate at the same speed. The power generated is proportional to the rate of combustion gas flow to the turbine and the temperature of the gas flow stream. Typically, many known CTGs have an operationally more dynamic behavior than STG (and their associated steam sources), therefore, CTGs may respond to system transients more rapidly.
One known method of maintaining a power reserve is to operate a CTG with at least one associated air guide vane and at least one fuel supply valve in a partially open, or throttled, position such that the CTG and the power grid are in an equilibrium, operating at some value less than the full rated load of the CTG. As discussed above for the STG, the difference between full load and the partial load is often referred to as spinning reserve. A controller senses a decrease in grid frequency and generates a signal that causes the air inlet guide vane and the fuel supply valve to open further within seconds of sensing the frequency transient. Since the compressor, the turbine and the generator are coupled to the same shaft, and since the generator that is synchronized to the grid decelerates as grid frequency is decreased, there exists an initial bias to channel less air into the CTG. This condition initiates a decreasing bias in CTG electric power generation that may negatively impact subsequent activities to increase CTG electric power generation. Furthermore, a bias to decrease air flow followed by a bias to increase air flow through the associated compressor may introduce a potential for a compressor surge, i.e., a substantially uncontrolled fluctuation of air flow and compressor discharge pressure, with surge potential being more pronounced at the lower end of compressor rated air flows. As the vane opens to increase the air flow and as the valve opens to increase the fuel flow, the mass flow rate of the combustion gas and the combustion gas temperature begin to increase within seconds of sensing the system frequency transient. Air and fuel are subsequently increased over time to facilitate establishing a modified equilibrium between the CTG and the power grid. In order to overcome the initial bias to decrease generation and then to accelerate the CTG, the combustion turbine may need to peak-fire, i.e., rapidly increase the rate of combustion to rapidly increase gas stream temperature while the subsequent increase of air flow follows. While the CTG may exhibit a more dynamic ability to respond to a frequency transient, many known CTGs may have temperature and temperature gradient limitations that may extend the time duration for increasing gas stream temperatures in order to mitigate stresses on a portion of the materials associated with the CTG. Otherwise, component stresses may increase and their associated life span may be negatively affected.
Many known steam generation apparatus and CTG are thermally most efficient operating in a range near the upper end of their operational power generation range. Maintaining a power generation level below that range may decrease thermal efficiency with a subsequent increase in cost of operation as well as possibly deny the owners of the facility potential revenue from the sale of the electric power held in reserve and routinely not generated.
Many known combined-cycle electric power generation facilities typically include at least one CTG and at least one STG. Some known configurations for such facilities include channeling the combustion gas exhaust from a CTG to a heat recovery steam generator (HRSG), wherein the thermal energy from the combustion gas exhaust boils water into steam, the steam subsequently being channeled to a STG. Typically, combined-cycle facilities are configured to use a CTG as the primary response mechanism for grid frequency transients while a STG is maintained as the secondary response. While this physical configuration offers benefits of efficiency and therefore economy of operation, the response configuration and method includes at least some of the aforementioned challenges in responding rapidly and effectively to a grid frequency transient.